美容护肤消费决策场景 AI 推荐行为解释性标注数据集

本数据集为基于美容护肤行业构建的人工智能问答推荐行为解释性标注数据集。数据通过主流人工智能问答系统对真实用户高频消费决策问题（如品牌选择、肤质匹配、预算/平价推荐、套装选择、长期使用安全性等）生成回答，并对回答结果进行结构化采集与整理而成。 数据以“问题—AI 输出答案—解释性标注字段”为核心结构：除保留原始问答文本外，还对每条样本补充标注问题主题、问题类型、价格敏感度、目标人群、肤质指向、核心诉求、答案结构类型、是否引用第三方来源及来源类型、推荐强度、是否风险提示、可疑营销品牌点名、品牌清单与风险点名品牌清单等字段，用于分析人工智能在不同消费语境下的品牌推荐倾向、来源引用特征、风险提示行为与稳定性。

This dataset is an interpretable annotated dataset for AI question answering and recommendation behavior, developed for the beauty and skincare industry. Data was collected by first using mainstream AI question answering systems to generate responses to high-frequency consumer decision-making questions from real users—such as brand selection, skin type matching, budget/affordable product recommendation, skincare kit selection, and long-term use safety—and then structurally collecting and organizing these response outputs. The core structure of the dataset follows the format of "Question – AI Generated Answer – Interpretable Annotation Fields": In addition to preserving the original question-answering text, each sample is supplemented with annotated fields including question topic, question type, price sensitivity, target demographic, skin type orientation, core user demands, answer structure type, whether third-party sources are cited and the specific source type, recommendation strength, presence of risk warnings, suspicious marketing brand mentions, full brand list, and risk-mention brand list. These annotation fields enable analysis of AI systems' brand recommendation tendencies, source citation characteristics, risk warning behaviors, and stability across different consumption contexts.